This cross-sectional study was undertaken to investigate the association between the reported frequency of consumption of vegetables and fruits, the choice of staple carbohydrate, and glucose intolerance. One thousand one hundred twenty-two subjects aged 40-64 years in a population-based study underwent an oral glucose tolerance test, and their food consumption was assessed using a food-frequency questionnaire. The crude prevalence of undiagnosed non-insulin-dependent diabetes mellitus (NIDDM) was 4.5%, and that of impaired glucose tolerance (IGT) 16.8%. The age-standardized prevalence rates were 2.3 and 11.2%, respectively. Frequent consumption of vegetables throughout the year was inversely associated with the risk of having NIDDM (odds ratio [OR] = 0.16; 95% confidence interval [CI] = 0.04-0.69). This association was maintained after adjustment for age, gender, and family history. Vegetable consumption during the summer months had a much weaker inverse association with the risk of having NIDDM that failed to reach statistical significance. A nonsignificant inverse association between frequent consumption of fruits and NIDDM was observed. Frequent self-reported pasta and rice consumption was associated with a reduction in the risk of having IGT and NIDDM. (OR = 0.62, 95% CI = 0.44-0.87, and OR = 0.51, 95% CI = 0.27-0.99, respectively) but this relationship was not independent of age. Whether these associations reflect specific effects of particular nutrients or are a reflection of the patterning of lifestyle factors remains to be determined.